I would like to extract data in batches from ScyllaDB and dump them into files or an OLAP database for data mining. Do you have a suggestion?
There are a number of ways to achieve this. One is using Presto (see this tech talk).
Another is by using Spark with ScyllaDB. By doing so, you deploy analytics workloads on information stored in ScyllaDB. You can learn more about this and see a hands-on lab in the Using Spark with ScyllaDB lesson on ScyllaDB University.